: Most T cell tolerance occurs during thymic development when potentially autoreactive T cells become eliminated. However, in order to prevent autoimmune disease, self-tolerance must also be established to peripherally expressed antigens not present in the thymus. This is an application to continue studies on the mechanism of tolerance within the CD8 T cell compartment to a transgene product, the influenza hemagglutinin (HA), expressed in the pancreatic islets. Tolerance in this model occurs as the result of abortive activation of naive T cells in the pancreatic lymph nodes. Experiments are proposed to define the antigen presenting cells (APC) involved in presentation of HA in the pancreatic lymph nodes and to determine the basis for inefficient peripheral tolerance in the neonatal period. To probe the mechanism of tolerance induction, mice that are genetically deficient in various TNF receptor family members (Fas, TNFR1, TNFR2 and CD30) that may signal peripheral deletion will be tested for development of tolerance to HA. Signals that may result in disruption of tolerance, through their activity on either the APC or the CD8 T cell, will be identified. The hypothesis will be tested that disrupting tolerance in the neonate may result in a repository of potentially autoimmune memory T cells. To determine if peripheral tolerance is important to prevention of autoimmune diabetes, tolerance to HA within the CD8 compartment will be evaluated in mice genetically predisposed to diabetes, NOD InsHA mice. If tolerance does not occur, the cellular and genetic mechanism responsible for this deficiency will be explored. Also, to extend the investigators' findings with HA to a naturally expressed beta cell antigen, the degree and mechanism of peripheral tolerance to GAD65 will be defined.